Evaluation of the internal construct validity of the Personal Care Participation Assessment and Resource Tool (PC-PART) using Rasch analysis

Evaluation of the internal construct validity of the Personal Care Participation Assessment and Resource Tool (PC-PART) using Rasch analysis

BMC Health Services Research, November 2014

25371103

Susan Darzins, Christine Imms, Marilyn Di Stefano, Nicholas F Taylor, Julie F Pallant

BackgroundThe Personal Care Participation Assessment and Resource Tool (PC-PART) is a 43-item, clinician-administered assessment, designed to identify patients¿ unmet needs (participation restrictions) in activities of daily living (ADL) required for community life. This information is important for identifying problems that need addressing to enable, for example, discharge from inpatient settings to community living. The objective of this study was to evaluate internal construct validity of the PC-PART using Rasch methods.MethodsFit to the Rasch model was evaluated for 41 PC-PART items, assessing threshold ordering, overall model fit, individual item fit, person fit, internal consistency, Differential Item Functioning (DIF), targeting of items and dimensionality. Data used in this research were taken from admission data from a randomised controlled trial conducted at two publically funded inpatient rehabilitation units in Melbourne, Australia, with 996 participants (63% women; mean age 74 years) and with various impairment types.ResultsPC-PART items assessed as one scale, and original PC-PART domains evaluated as separate scales, demonstrated poor fit to the Rasch model. Adequate fit to the Rasch model was achieved in two newly formed PC-PART scales: Self-Care (16 items) and Domestic Life (14 items). Both scales were unidimensional, had acceptable internal consistency (PSI =0.85, 0.76, respectively) and well-targeted items.ConclusionsRasch analysis did not support conventional summation of all PC-PART item scores to create a total score. However, internal construct validity of the newly formed PC-PART scales, Self-Care and Domestic Life, was supported. Their Rasch-derived scores provided interval-level measurement enabling summation of scores to form a total score on each scale. These scales may assist clinicians, managers and researchers in rehabilitation settings to assess and measure changes in ADL participation restrictions relevant to community living.Trial registrationData used in this research were gathered during a registered randomised controlled trial: Australian and New Zealand Clinical Trials Registry ACTRN12609000973213. Ethics committee approval was gained for secondary analysis of data for this study.